<p>"How much of the class at UMichigan, Northwestern or Cornell are in those more obscure majors? Taking Cornell, that I'm more familar with, there's no way more than 1,000 of 3,565 (which seems high to me) of the class are in hotel management or CALS (some of whom will end up pursuing an MBA eventually). However, even throwing out 1,000 of the class out of the calculation, Cornell is still not going to come out better than around where Haverford or Bowdoin comes out ~ 4% of the class."</p>
<p>Actually Gellino, 1,000 out of Cornell's 3,565 graduating class sounds about right. Add to that 1,000 another 1,000 who belong to the colleges of Art/Architecture or Human Ecology and you get more than 50% of Cornell's student body that is simply not likely to even want to apply to Medical, Law or Business graduate programs. That's 2,000 out of 3,600. From my personal observations, the ratio of undergrads interested in those three professional graduate programs at colleges like Dartmouth, Duke or Georgetown and several LACs such as Williams, outnumbers that at more diversified or less pre-professional undergraduate colleges such as Caltech, Chicago, Cornell and Northwestern by a ratio of 2:1. In the case of the elite publics, I would say they outnumber them 3:1.</p>
<p>"Most of the top grad schools are on the East coast. If anything gets shafted in this study, it's the West coast; not the Midwest. Do you really think this survey is going to come out materially different if for MBA programs Stanford, Kellogg, Columbia, UMichigan, Darden are also used and for JD programs, Stanford, NYU, Cal/Berkeley, UVA, UMichigan are also used. I certainly doubt the results would change very much."</p>
<p>I beg to differ. If the survey replaced Sloan with Kellogg and Tuck with Ross and HBS with Stanford Business school and Columbia Law with Stanford Law and Columbia Medical with Michigan Medical, I guarantee you would get a significantly different picture. Why even replace programs. If instead of sticking to the "top 5", the WSJ had surveyed "the top 10" the results would have been more even.</p>